|
|
||||||||
a Dep. of Plant Breeding, 252 Emerson Hall, Cornell Univ., Ithaca, NY 14853
b ICARDA, CIMMYT/ICARDA, Durum Improvement Program, P.O. Box 5466, Aleppo, Syria
* Corresponding author (mes12{at}cornell.edu)
| ABSTRACT |
|---|
|
|
|---|
Abbreviations: MITE, miniature inverted-repeat transposable element
| INTRODUCTION |
|---|
|
|
|---|
Lipoxygenases form a family of enzymes that catalyze the breakdown of polyunsaturated fatty acids in plants, animals, and microorganisms (Prigge et al., 1996). In plants, lipoxygenases are found in seeds, seedlings, and leaves. The isoenzymes in each of these tissues are distinct from each other. In higher plants, lipoxygenase reacts with the substrate linoleic acid or linolenic acid and forms hydroperoxides. These products are thought to be involved in plant defense, wound response, senescence, and development (Hildebrand, 1989). Under certain processing conditions, high levels of lipoxygenase activity destroy the yellow color in pasta by oxidation (Joppa and Williams, 1988). Some plant lipoxygenase reaction products also are implicated in the production of aroma or undesirable flavors and odors. In barley (Hordeum vulgare L.), this enzyme is thought to be responsible for staling in beer (Drost et al., 1990; Wu et al., 1997). Lipoxygenase activity is directly affected by storage conditions. Kaukovirta et al. (1998) measured oxidation of linoleic acid in flour suspensions of barley and malt samples. Results showed that there was great variability of lipoxygenase activity profiles from a single malting, depending on the duration of storage before the assays. They concluded that the rate of lipoxygenase reaction should be considered as a quality factor of malt because of its effect on malting. In soybean [Glycine max (L.) Merr.] seeds, lipoxygenases are responsible for production of the unpleasant beany flavors that have limited the development of soybean-protein products for human consumption (Shibata, 1996). The hydroperoxides that are formed by breaking down the unsaturated fatty acids are cleaved by hydroperoxide lyase to form aldehydes, which result in the off-flavor (Shibata, 1996).
One of the primary quality traits important for durum products is a bright yellow color. The pigments responsible for this color are luteins, a class of carotenoids (Laignelet, 1983). Besides the genetic potential for production of these pigments in the seeds, a number of other factors affect the color of the end product, such as pigment breakdown during grain storage, the extraction rate from milling, and degradation during processing (Irvine, 1971; McDonald, 1979). Lipoxygenases are reported to be among the more important enzymes that contribute to loss of yellow color in durum seeds and products (Borrelli et al., 1999; Irvine and Anderson, 1953; Manna et al., 1998).
Borrelli et al. (1999) found that pasta processing caused a 16% loss in beta-carotene content versus 8% loss from milling. Lipoxygenase activity (pH 4.86.6) in the semolina was the main factor causing the color loss. Manna et al. (1998) compared mRNA levels for lipoxygenase activity in nine durum cultivars with lipoxygenase activity (measured at pH 4.8 and 10.2) and beta carotene levels. They found high negative correlations (r > -0.95) between lipoxygenase activity measured at pH 10.2 and beta carotene levels. In contrast to Borrelli et al. (1999), they did not find a correlation between lipoxygenase activity at pH 4.8 and beta carotene content in the semolina.
Reducing lipoxygenase activity in varieties possessing other high quality attributes is highly desirable to maintain fresh flavor and yellow pasta color. Locating genes controlling lipoxygenase activity in grains would facilitate breeding efforts to select low lipoxygenase genotypes. Mutants lacking lipoxygenase in soybean seeds have been isolated (Furuta et al., 1996; Kitamura, 1993; Lambrecht et al., 1996; Narvel et al., 1998). Narvel et al. (1998) determined the effects of genetically eliminating lipoxygenase isoenzymes in soybean. Successful backcrosses were made with 3 lipoxygenase-null alleles. Their work showed no differences between normal and lipoxygenase-null lines for seed yield, maturity, lodging resistance, and seed protein content, suggesting that it is possible to develop lipoxygenase-free soybean cultivars with appealing agronomic characteristics. Lipoxygenase isoenzymes already have been separated in barley (Wu et al., 1997). Research on the expression of lipoxygenase in barley embryos during germination suggested that the lipoxygenase-1 isoenzyme was responsible for the lipoxygenase activity in sound seeds (Holtman et al., 1996). In addition, three lipoxygenase cDNA sequences (LoxA, LoxB, LoxC) have been characterized in barley, with LoxA corresponding to the lipoxygenase1 isozyme, and LoxC corresponding to lipoxygenase2 (van Mechelen et al., 1999, 1995)
The purpose of this study was to map the loci associated with lipoxygenase activity in seeds of a durum wheat population and to determine if lipoxygenase activity is associated with flour color. Lipoxygenase activity was measured in seed because it is the lipoxygenase activity in ungerminated seed that affects semolina quality in durum wheat. By identifying the correlation between the phenotypic variation for this trait and the variation at the DNA sequence level, a more rapid assessment of lipoxygenase activity may be possible.
| MATERIALS AND METHODS |
|---|
|
|
|---|
![]() |
![]() |
The PCR product from each parental line was used in TA cloning (Invitrogen, Carlsbad, CA) to clone into the pCR 2.1 vector and were transformed into competent Escherichia coli cells. Isolated plasmid DNA from 7 Jennah Khetifa colonies and 6 Cham 1 colonies were sequenced on an automated sequencer at the BioResource Center at Cornell University.
For each parent, two unique sequences were amplified (see Fig. 1) , one from the A and one from the B genome as determined by linkage tests. The wheat gene homologous to LoxA designated Lpx (MacIntosh, 1988) was identified on the basis of the published barley LoxA sequence (van Mechelen et al., 1995). The two DNA sequences within each parent were 95% identical and the sequences from the same genome between parents were 99% identical. The PCR products from these primers were separated on 0.9% (w/v) agarose LE gels and scored for parental type on the Jennah Khetifa/Cham 1 population of 113 lines.
|
|
Trait Evaluation
Water-saturated n-butanol was used to extract the pigment from ground wheat samples (Williams et al., 1988). The filtered extract was read at 440 nm on a spectrophotometer and converted to ppm beta carotene, on the basis of a standard graph of synthetic beta-carotene. Pigment in ground samples was also estimated by near-infrared reflectance spectroscopy (NIR), according to the methods used by ICARDA (Williams et al., 1988).
Percent vitreous seeds was measured by counting the number of vitreous kernels out of a sample of 100 (Williams et al., 1988). Seed weight was determined by weighing 200 kernels to the nearest 0.01 g (Williams et al., 1988). Protein content was measured by NIR (Williams et al., 1988) using 20 g of ground seeds.
Enzyme Extraction and Assay
Using the parents and 113 RI lines from the Jennah Khetifa/Cham 1 cross, we soaked 0.5 g of finely ground seed in 3 mL 0.1 M potassium phosphate buffer, pH 6.0, for 1 h according to the method of Wu et al. (1997) to obtain flour extracts. The flour extracts were then centrifuged for 10 min at 12000 g at 4°C and filtered with a syringe and disc filter, 0.2-µm pore size. Extracts were immediately frozen at -80°C until assays were performed.
Linoleic acid substrate (8 mM) was prepared according to the method of Wu et al. (1997), which was a modification of the method of Surrey (1964). Linoleic acid is highly reactive in the presence of oxygen. All components of the substrate were thoroughly flushed with nitrogen gas and the linoleic acid was kept in an atmosphere of nitrogen in a glove box. Aliquots were frozen at -80°C until use. The same batch of substrate was used with all samples to ensure the same level of autooxidation. Each sample was assayed in duplicate.
Lipoxygenase activity was measured by spectrophotometric methods described by Wu et al. (1997). Lipoxygenase activity was measured by the increase in absorbance at 234 nm, which is a measurement of the amount of hydroperoxide that was formed from the reaction. The reaction was initiated by combining 0.04 mL of thawed substrate with 0.92 mL 0.1 M potassium phosphate buffer (pH 6.0) and adding 0.04 mL of the flour extract. Changes in absorbance over a one minute time period were recorded for each line in duplicate and parent lines were tested four times.
Data Analysis
The molecular marker map for the Jennah Khetifa/Cham 1 population (Nachit et al., 2001) consists of 306 markers spaced an average of 12 centimorgans apart. The length polymorphism between the parents from the Lpx primers was mapped by means of the "Try" and "Ripple" commands in the MapMaker computer program (Lander et al., 1987). Phenotypic data consisting of means from repeated samples were entered in QGene computer program format, along with data for flour color, protein content, vitreous appearance, kernel weight, pigment, and kernel color. Single point analysis followed by interval regression analysis were performed by the computer program QGene (Nelson, 1997) to identify loci associated with variation in the traits. A threshold of LOD 3 in this recombinant inbred population yields a significance level of 0.05 (Lander and Botstein, 1989).
| RESULTS AND DISCUSSION |
|---|
|
|
|---|
Although the point mutation differences in the coding regions, most of which are silent mutations, provide some information as to the relatedness of the amplified sequences, it is the intron region that shows the most polymorphism. Three different intron variations were found, resulting from insertiondeletion events caused by a MITE of the Stowawayclass. One of the group of Cham 1 sequences (e.g., C7.1.2) has all five characteristics of Stowaway elements: a conserved 11-bp terminal inverted repeat, small size (80323 bp), AT rich, strong target site preference for TA, and the potential to form DNA secondary structures (Bureau and Wessler, 1994). The Cham 1 element has high sequence similarity to three other elements located near Triticeae genes, two of which have been classified as Stowaway-type MITEs (Bureau and Wessler, 1994; Peterson and Seberg, 2000) (Fig. 2) . In C7.1.2, the complete element is followed by a direct repeat of part of the second half of the element (Fig. 1 and 2). In the other Cham 1 sequence (C5.36.2), the element is missing. In one of the Jennah Khetifa sequences (J4.2), only the first half of the element and the last 11-bp inverted repeat is present, while in J2.2 the entire element is missing (Fig. 2). The 83-bp intron amplified from the barley cultivars Steptoe and Morex did not contain the Stowaway element (data not shown).
The polymorphism seen both within and between the durum wheat varieties Jennah Khetifa and Cham 1 at this site indicates potential instability caused by the MITE insertion. MITEs preferentially insert in gene-rich regions, and may affect gene expression by inserting in cis-regulatory regions of genes, whether in the 3' region or in introns. Further studies are needed to discern whether different MITE families, or MITEs in different species, may have different degrees of polymorphism. If MITEs are indeed highly polymorphic loci in gene-rich regions, they may provide the basis for an efficient marker system in species with high MITE copy numbers, as has been suggested for use in maize (Casa et al., 2000). These results for the Lpx-B1 locus in durum wheat provide an example of the usefulness of a MITE polymorphism in a simple PCR-based marker system for differentiating alleles.
Trait Means and Ranges
Mean lipoxygenase activity, as measured by change in units of spectrophotometric absorbance over a 1-min period, was 0.752 ± 0.066 for Jennah Khetifa versus 0.284 ± 0.022 for Cham 1. Lipoxygenase activity of lines in the population ranged from 0.1875 to 1.3901. Samples were used from multiple environments for the parental lines, which showed no significant difference between environments (variance for Jennah Khetifa was 0.0006 and 0.0007 for Cham 1). Therefore, only one replication was needed, in duplicate, for the population sample assays. Flour color measured by NIR ranged from 8.2 to 4.4 ppm among the lines with parental values of 5.2 for Jennah Khetifa and 6.1 for Cham 1. Yellow pigment extraction yielded 5.3 for Cham 1 and 4.6 for Jennah Khetifa with the population ranging from 7.1 to 3.4. The range for vitreous seeds was 89 to 100% for the population and 96% for Jennah Khetifa and 98% for Cham 1. The range for protein content was 12.8 to 17.4%. Seed weight had a range from 22.9 to 39.9.
QTL Analyses
The polymorphism detected by the lipoxygenase primers mapped to the short arm of wheat chromosome 4B on the basis of the previously published map (Nachit et al., 2001). Hart and Langston (1977) mapped zone 1 lypoxygenase isozymes to chromosome arms 4AS, 4BL, and 4DS of hexaploid wheat. In the original report, Lpx-B1 was located on the long arm and Lpx-A1 on the short arm. However, that same year, wheat geneticists reversed the designations of those chromosomes so the chromosome location of the lypoxygenase locus reported here is consistent with the location of the previously mapped isozyme. Quantitative trait locus analysis showed that lipoxygenase activity was significantly (LOD = 9.8) associated with this Lpx locus on the short arm of chromosome 4B (Table 1) . This locus was responsible for 36% of the variation in lipoxygenase activity in this population. Additional loci, Xutv135c, Xbcd327, XPacMcag8, and XPaggMcag5, on chromosome 4B were associated with enzyme activity at a lower level because they are linked to Lpx. A second locus, Xrz444 on chromosome 2B, accounted for 9% of the variation in lipoxygenase activity; however, interval analysis mapped it at LOD = 2.11, well below the significance threshold of 3.0. No other loci in this population were significant for lipoxygenase activity.
|
Percent vitreous seeds was associated with the Lpx locus, accounting for 10% of the variation with a LOD score of 2.74, just slightly below the significance threshold. Jennah Khetifa was the high parent for this locus for both lipoxygenase activity and percent vitreous kernels. These results suggest that percent vitreous kernels may be influenced by Lpx-B1 or a closely linked gene.
Thousand-seed weight also was measured to determine if this trait might be associated with lipoxygenase activity or flour color. One major locus on chromosome 5A accounted for 20% of the variation with a LOD score of 5.78. This is the same locus that was significant for flour pigment but the high parents are reversed. This indicates that this locus is associated with larger kernels and less pigment possibly due to a dilution effect of the starch. The embryo and seed coat have more lipoxygenase than the endosperm. It would seem that smaller kernels would produce more lipoxygenase activity in the assay used in this study, because there are more embryos and seed coats per unit of weight from smaller seeds. However, Cham 1 had both smaller kernels and less lipoxygenase activity. The enzymatic activity over one minute for a 0.5-g sample was used as a measure of lipoxygenase activity, so it is not possible to determine if the lipoxygenase gene from Cham 1 encodes a less efficient form of the enzyme or if the locus is simply expressed at a lower level, producing less enzyme. Future studies should evaluate other lipoxygenase isozymes to determine if they are responsible for variation in kernel pigment in this population.
In this study, allele specific primers were designed that can be used in a marker assisted selection program to reduce lipoxygenase levels in durum wheat using the direct allele selection method (Sorrells and Wilson, 1997). The determination that natural allelic variation at the Lpx-B1 locus in durum wheat is associated with different lipoxygenase levels represents the first step towards implementation of marker assisted selection for low-lipoxygenase activity.
| ACKNOWLEDGMENTS |
|---|
Received for publication July 18, 2001.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
F. Breseghello, P. L. Finney, C. Gaines, L. Andrews, J. Tanaka, G. Penner, and M. E. Sorrells Genetic Loci Related to Kernel Quality Differences between a Soft and a Hard Wheat Cultivar Crop Sci., August 1, 2005; 45(5): 1685 - 1695. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Vadose Zone Journal | |||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||